Association for Public Policy Analysis & Management

Since the passage of federal legislation in 1975 guaranteeing a free public education for all students with disabilities (now titled Individuals with Disabilities Education Act or IDEA), special education enrollment in public schools has grown from 3.6 million (8% of students) to nearly 6.6 million (13%), with some cities such as New York and Boston serving over 20% (NCES, 2015). Although the provision of free public schooling and appropriate services for students with disabilities (SWDs) is rarely contested by policymakers or the public, there is persistent literature documenting and debating disproportionate representation of black and minority students, and worry that they neither exit services in timely ways nor adequately benefit academically, but rather are warehoused and stigmatized (Coutinho and Oswald, 2005; Morgan et al., 2010, 2012, 2015, 2017; Oswald et al., 2001, p. 361; Schiefter, 2011; 2014; Skiba et al., 2005, p. 142).

Concerns about disproportionality are not simply academic, but have led to changes in law, with Congress mandating data collection on disproportionality, placement into special education, and type of disciplinary actions. If significant disproportionality is found, three remedial provisions are triggered (Albrecht et al., 2012)

Existing research has explored various reasons for why disproportionality might exist, including socioeconomic factors (Brooks-Gunn and Duncan, 1997; Engle and Black, 2008; Nelson and Sheridan, 2011; Skiba, 2008), biased teacher referrals (Skiba et al., 2008), and accountability incentives (Cohen, 1997; Figlio and Getzler, 2006). Among these factors, however, the role of schools themselves is rarely cited. Yet school context may be important to the understanding of racial representation in special education.

In this paper, we develop hypotheses about how schools might affect the representation of students by race and ethnicity in special education and then examine empirical evidence using data in a large urban district, New York City (NYC), with more than 1700 schools and 1.1 million students, over 20% of whom are students with disabilities (SWDs). Using logistic regression models to describe patterns, several notable findings emerge in our results. Race/ethnicity gaps in SWD classification differ by school organization levels, where white, black, and Hispanic students have more similar rates of SWD classification in elementary school, but white student classification drops in middle school while black and Hispanic classification increases, creating larger black-white and Hispanic-white gaps. School level characteristics also significantly predict differences between races and ethnicities in SWD classification rates – students’ own race and teacher race concentrations reduce gaps compared to whites; school climate (safety and discipline) reduce differential placement of students into special education; and large school size reduces gaps, but the effect is very small.

Our goal is to understand such school patterns in greater detail, providing policymakers, practitioners, and researchers with a better way of understanding and addressing factors contributing to disproportionality in special education.